Abstract

Recovering a matrix from a sampling of its entries is a problem of rapidly growing interest and has been studied under the name of matrix completion. It occurs in many areas of engineering and applied science. In most machine learning and data mining applications, it is possible to leverage the expertise of human oracles to improve the performance of the system. It is therefore natural to extend this idea of "human-in-the-loop" to the matrix completion problem. However, considering the enormity of data in the modern era, manually completing all the entries in a matrix will be an expensive process in terms of time, labor and human expertise, human oracles can only provide selective supervision to guide the solution process. Thus, appropriately identifying a subset of missing entries (for manual annotation) in an incomplete matrix is of paramount practical importance, this can potentially lead to better reconstructions of the incomplete matrix with minimal human effort. In this paper, we propose novel algorithms to address this issue. Since the query locations are actively selected by the algorithms, we refer to these methods as active matrix completion algorithms. The proposed techniques are generic and the same frameworks can be used in a wide variety of applications including recommendation systems, transductive/multi-label active learning, active learning in regression and active feature acquisition among others. Our extensive empirical analysis on several challenging real-world datasets certify the merit and versatility of the proposed frameworks in efficiently exploiting human intelligence in data mining/machine learning applications.

Original languageEnglish (US)
Article number6729492
Pages (from-to)81-90
Number of pages10
JournalUnknown Journal
DOIs
StatePublished - 2013

Fingerprint

Problem-Based Learning
Data Mining
Data mining
Learning systems
Intelligence
Recommender systems
Names
Labels
Personnel
Sampling
Machine Learning
Datasets

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Chakraborty, S., Zhou, J., Balasubramanian, V., Panchanathan, S., Davidson, I., & Ye, J. (2013). Active matrix completion. Unknown Journal, 81-90. [6729492]. https://doi.org/10.1109/ICDM.2013.69

Active matrix completion. / Chakraborty, Shayok; Zhou, Jiayu; Balasubramanian, Vineeth; Panchanathan, Sethuraman; Davidson, Ian; Ye, Jieping.

In: Unknown Journal, 2013, p. 81-90.

Research output: Contribution to journalArticle

Chakraborty, S, Zhou, J, Balasubramanian, V, Panchanathan, S, Davidson, I & Ye, J 2013, 'Active matrix completion', Unknown Journal, pp. 81-90. https://doi.org/10.1109/ICDM.2013.69
Chakraborty S, Zhou J, Balasubramanian V, Panchanathan S, Davidson I, Ye J. Active matrix completion. Unknown Journal. 2013;81-90. 6729492. https://doi.org/10.1109/ICDM.2013.69
Chakraborty, Shayok ; Zhou, Jiayu ; Balasubramanian, Vineeth ; Panchanathan, Sethuraman ; Davidson, Ian ; Ye, Jieping. / Active matrix completion. In: Unknown Journal. 2013 ; pp. 81-90.
@article{4aeb8d32a97c4b99ad9f97b3fe92e1db,
title = "Active matrix completion",
abstract = "Recovering a matrix from a sampling of its entries is a problem of rapidly growing interest and has been studied under the name of matrix completion. It occurs in many areas of engineering and applied science. In most machine learning and data mining applications, it is possible to leverage the expertise of human oracles to improve the performance of the system. It is therefore natural to extend this idea of {"}human-in-the-loop{"} to the matrix completion problem. However, considering the enormity of data in the modern era, manually completing all the entries in a matrix will be an expensive process in terms of time, labor and human expertise, human oracles can only provide selective supervision to guide the solution process. Thus, appropriately identifying a subset of missing entries (for manual annotation) in an incomplete matrix is of paramount practical importance, this can potentially lead to better reconstructions of the incomplete matrix with minimal human effort. In this paper, we propose novel algorithms to address this issue. Since the query locations are actively selected by the algorithms, we refer to these methods as active matrix completion algorithms. The proposed techniques are generic and the same frameworks can be used in a wide variety of applications including recommendation systems, transductive/multi-label active learning, active learning in regression and active feature acquisition among others. Our extensive empirical analysis on several challenging real-world datasets certify the merit and versatility of the proposed frameworks in efficiently exploiting human intelligence in data mining/machine learning applications.",
author = "Shayok Chakraborty and Jiayu Zhou and Vineeth Balasubramanian and Sethuraman Panchanathan and Ian Davidson and Jieping Ye",
year = "2013",
doi = "10.1109/ICDM.2013.69",
language = "English (US)",
pages = "81--90",
journal = "Scanning Electron Microscopy",
issn = "0586-5581",
publisher = "Scanning Microscopy International",

}

TY - JOUR

T1 - Active matrix completion

AU - Chakraborty, Shayok

AU - Zhou, Jiayu

AU - Balasubramanian, Vineeth

AU - Panchanathan, Sethuraman

AU - Davidson, Ian

AU - Ye, Jieping

PY - 2013

Y1 - 2013

N2 - Recovering a matrix from a sampling of its entries is a problem of rapidly growing interest and has been studied under the name of matrix completion. It occurs in many areas of engineering and applied science. In most machine learning and data mining applications, it is possible to leverage the expertise of human oracles to improve the performance of the system. It is therefore natural to extend this idea of "human-in-the-loop" to the matrix completion problem. However, considering the enormity of data in the modern era, manually completing all the entries in a matrix will be an expensive process in terms of time, labor and human expertise, human oracles can only provide selective supervision to guide the solution process. Thus, appropriately identifying a subset of missing entries (for manual annotation) in an incomplete matrix is of paramount practical importance, this can potentially lead to better reconstructions of the incomplete matrix with minimal human effort. In this paper, we propose novel algorithms to address this issue. Since the query locations are actively selected by the algorithms, we refer to these methods as active matrix completion algorithms. The proposed techniques are generic and the same frameworks can be used in a wide variety of applications including recommendation systems, transductive/multi-label active learning, active learning in regression and active feature acquisition among others. Our extensive empirical analysis on several challenging real-world datasets certify the merit and versatility of the proposed frameworks in efficiently exploiting human intelligence in data mining/machine learning applications.

AB - Recovering a matrix from a sampling of its entries is a problem of rapidly growing interest and has been studied under the name of matrix completion. It occurs in many areas of engineering and applied science. In most machine learning and data mining applications, it is possible to leverage the expertise of human oracles to improve the performance of the system. It is therefore natural to extend this idea of "human-in-the-loop" to the matrix completion problem. However, considering the enormity of data in the modern era, manually completing all the entries in a matrix will be an expensive process in terms of time, labor and human expertise, human oracles can only provide selective supervision to guide the solution process. Thus, appropriately identifying a subset of missing entries (for manual annotation) in an incomplete matrix is of paramount practical importance, this can potentially lead to better reconstructions of the incomplete matrix with minimal human effort. In this paper, we propose novel algorithms to address this issue. Since the query locations are actively selected by the algorithms, we refer to these methods as active matrix completion algorithms. The proposed techniques are generic and the same frameworks can be used in a wide variety of applications including recommendation systems, transductive/multi-label active learning, active learning in regression and active feature acquisition among others. Our extensive empirical analysis on several challenging real-world datasets certify the merit and versatility of the proposed frameworks in efficiently exploiting human intelligence in data mining/machine learning applications.

UR - http://www.scopus.com/inward/record.url?scp=84894666258&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84894666258&partnerID=8YFLogxK

U2 - 10.1109/ICDM.2013.69

DO - 10.1109/ICDM.2013.69

M3 - Article

SP - 81

EP - 90

JO - Scanning Electron Microscopy

JF - Scanning Electron Microscopy

SN - 0586-5581

M1 - 6729492

ER -